Machine learning in trading: theory, models, practice and algo-trading - page 2567
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
I am not interested in the cotier itself. I am interested in the predictor's ability to predict the teacher. To me the biggest mistake of the vast majority of traders is in their attempts to solve the problems of the kotir itself. And we need the prediction of the teacher. That's a completely different problem.
If the predictors and answers are clearly formalized, there is probably no need for a price model. If they are searched for and constructed, then a price model would not hurt.
If the predictors and answers are clearly formalized, there is probably no need for a price model. If they are being searched for and constructed, then a price model would not hurt.
The future market model is always formed from the past. The past forms the future model depending on the past and the current state of the world economy and politics.
The current state of the economy is the first factor in the direction of the currency component. There is no doubt about the influence of the words of the First Worlders on the exchange rates.
It looks beautiful in the tester when the data is ready.
P.s. What's the point of this whole post?
One of the most important indicators of the impact on the currency market is news that we haven't heard about yet. Historical data + current news will form a new pattern, but it will be different from the one assumed in MO.
I.e. the cul-de-sac will always be in a different place.
https://youtu.be/KxDZZjM6ltI
Well, if he's going into antiquity, he should have started with Aristotle and his concept of essence.
Well, if he is going back in time, he should have started with Aristotle and his concept of essence.
Did you watch the whole thing?
Almost. Definition is what he calls "nominal definition". There is also "real definition" or essence according to Aristotle.
Almost. By definition he means what is commonly called "nominal definition. There is also "real definition" or essence definition according to Aristotle.
Is that really the most important thing you've picked out for yourself from the whole lecture?
What in reasoning can be more important than their starting point?)
The concepts of context and semantics have long been known, as well as their complexity and multilayeredness.
Regarding MO algorithms - it seems to boil down to just another complication in the architecture of networks that are already complex enough to achieve retraining)
The interpretations of mythology are not uninteresting, but there's nothing original about that either.
Regarding MO algorithms - apparently it all comes down to just another complication in network architecture
In fact, if you look at the problem fundamentally, the answer becomes much simpler.
If you have input data that has anything to do with the target, any NS machine will do the job. And the more accurate the input data describe the target, the better the results of network training, the longer its lifetime with constant quality, etc.
That is, the essence of work in the field of MI is not in endless search of NS architectures, training methods and other stuff, but in SELECTING quite working AI and endless search of input data for even better results of training and work of the model as a whole.
This is what is at work in machine learning. Choosing an artificial intelligence system (not just a NS, but an AI) and working with the chosen system to find the best inputs for the specified target function. Sometimes some data works, sometimes others built on diametric transformations, but half a year the difference works, half a year the averaging works and you need to be able to adjust to that alas.
Otherwise: If there is no informative relationship between input and output data in the system, then no AI system can build a describing law between input and output and therefore all proposed models will have random operation or not, since there is no REAL law between input and output in principle!